MarkDownload - Markdown Web Clipper

The clipper is working well on Microsoft Edge. Thanks so much for this!

This is a minor quibble but it seems to place all the attachments in my Downloads folder, regardless of where I choose to put the note.

If the browser is set to “ask me where to download each file” then you must manually select where to download the (potentially dozens of) image files. If you have the browser set to automatically download to your downloads folder, but the extension set to select a folder, then the extension will auto-populate the attachments folder into the folder selection window (which is set to Downloads) and place all the images there.

It’s easy enough to move these attachments to the proper place however.

Thank you for this web clipper. I have tried many other ones for Notion, Evernote and others and was utterly disappointed. This is particularly for the Show Notes from a podcast by Dr. Peter Attia (The Drive). Every other web clipper would only download up to the free preview part no matter whether I am logged on to the Members site.

Markdownload worked like a charm.

Many thanks!

– Sai

Edit it is working now. Wasn’t before restart.

Wonderful web clipper, and I’ve tried many of them.

One question, though: where can I find a list of all supported meta tags? I am trying to achieve something, but some meta tags that I discovered online weren’t supported, some don’t work as expected.

What I’m trying to achieve here is extract the top level domain of a website so that I can create a folder where I will keep all the images from that website. Something like, so all images from BBC will be kept under that same folder automagically.

Is this possible? baseURI extracts the whole url.

I love this extension!

I have one want - tag completion.

I like to to add tags each time I clip a web page. That helps me find those clipped pages later.

I can add tags with this extension, but it’s totally free-form - there’s no connection with my existing obsidian tags. I’d like it to show me a list of my obsidian tags as soon as I type # and then narrow down that list of tags as I continue to type.

In this example below I typed #bookmark but there’s nothing to guide me away from typing #bookmarks. And I’ll end up with many near-duplicate bookmarks that will be a mess.

I’m guessing that this would be difficult for a browser extension alone, and some sort of Obsidian plugin would be needed?

Any alternatives or ideas to help me get where I want to go?

1 Like

MarkDownload files in your -Inbox

For technical reasons, the MarkDownload browser add-on can only save to the browser’s Downloads folder, which is usually ~/Downloads. We usually let it save downloaded pages into the MarkDownload subfolder.

Now if you want to have direct access to the downloaded files in Obsidian, you can symlink the ~/Downloads/MarkDownload folder (and its subfolders, in case you also save the images from webpages) into your Obsidian vault’s -Inbox folder.

In my case, the Obsidian -Inbox is at ~/Dokumente/Obsidian/Knowledgebase/-Inbox, so the commands are as follows (in a Linux terminal—use your folder names!):

cd ~/Dokumente/Obsidian/Knowledgebase/-Inbox
ln -s ~/Downloads/MarkDownload/

and presto! you have a Markdownload subfolder in your Inbox and can directly access the downloaded pages, edit them and put them into their final place in Obsidian.

View in the file manager

View in Obsidian

Note: This should also work on MacOS and Windows, you might have to use other commands, though (like mklink on Windows).

Note 2: If you are syncing your vault between machines (using the same folder structure), the MarkDownload folder in your machine’s Obsidian will only ever show what’s in that machine’s MarkDownload folder. Remember we’re only syncing a symlink, not the real ~/Downloads/MarkDownload folder (which of course isn’t synced, and thus different on each machine).

So it pays reworking the downloaded files soon and eventually putting them into their final place in your vault (where they then will sync).

I thought I’d finally share this, since colleagues on the Discord assure me Obsidian Sync can handle symlinks correctly. Syncthing also does.

So it should work even if you sync your vault, only of course requiring the same folder structure. Syncing to my Android breaks the symlink there (no ~/Downloads/MarkDownload folder on Android) but since the link just gets copied as a symlink, it doesn’t matter.

@Licat also tells us we shouldn’t symlink across filesystems (like to another mounted folder that’s actually on a NAS or so), because Obsidian’s internal file reader can’t cope with that.


Thanks for the tip.

I think a complete solution would require inotify to detect file changes in the folder and sync them with the vault.

Anyways, great extension, OP. Basically replaced pocket for me.

Btw, is it possible to include the link/url of the original webpage on the clipped markdown? Like on the YAML front matter or just normal obsidian link in [ ]. I may have missed the option.

Could be useful when I need to trace back the original webpage.
I basically use it as a read-it-later clipper like pocket.

Thanks again for the very useful extensions.

Put something like this in as part of your frontmatter template

articleweblink: "[HERE]({baseURI})"

1 Like

Oh so that’s what “Include front/back template” option does. Thanks a lot.

1 Like

Pleasure - It’s a great plugin and very tweakable - I have it setup to put images into a folder in the same fashion as I have obsidian configured so I just drag and drop into my vault and presto it works

1 Like

You can include the source URL in the front matter and in the body text. For a while I had it in source: in the yaml front matter and then "Source: " in the body text itself but this felt duplicative.

1 Like

Yes, but only hard links work, I’m using bash shell. The usual soft links provided by the “Finder” don’t do the job. At least not on Mojave.

Thanks for the tip!

Thanks for testing this! Hard links can be dangerous … and Finder aliases aren’t real symlinks. But you can use real symlinks on Mac, see here → Ability to index / add external folders from across computer - #6 by Moonbase59

I don’t understand why it should be dangerous to use hard links, but ln - s works perfectly, Thanks so much!

Someone should pin your help.
Because it’s highly unproductive to copy only originals into Obsidians “database”

is there a way to get the website address? (the first part of url before the first /) it is a useful metadata

i think there was more custom text at the beginning of development of this extension than it is now. is it my memory or really there is less option now?

1 Like

FWIW: I’m on a Mac and I discovered this wonderful Firefox add-on. I too noticed that it would only save to the Downloads folder. I made a soft link from my bookmarks folder in my vault to a “bookmarks” folder inside the ~/Downloads folder. Works perfectly fine.

Every time I make a soft link I always cd to where I wish it to be, to avoid typing out long location strings multiple times (risk for typos). As the format is “ln -s [full-path-to-original-folder/file] [location-of-link]” I only need to end the command with the name for the file folder at the location of the copy, and I don’t mix up if the original or link comes first/last.

So, in a terminal (iTerm2 in my case) window I did:

cd ~/Downloads 
ln -s /Users/mathiashellquist/git/resources/bookmarks/ bookmarks

Job done. Works great. As I sync my vaults via Git I don’t have to care about the ~/Downloads folder, which is where the symbolic link lives, syncing or not. I will only be syncing “real” files via Git to all my devices, and the path within the vault is the same on all my used platforms (macOS, Linux, Win, iOS).

PS. As I was scrolling this thread today when I found it, my take on Chrome is that it should be avoided at all cost. It is basically spyware. If you care about Privacy in the slightest, and you should, you shouldn’t use Chrome. There are also good Blink (Chromes “rendering engine”) based alternatives out there, that doesn’t collect all your personal details. There is no need to hand over all your personal details to Google these days. IMHO. :slight_smile:


You can clean up any similarly named tags using the rename function of tag wrangler plugin!

I love the markdownload metadata extraction. But have 2 questions.

  1. I usually don’t want to extract the body of the website / article. Is there an option to by default NOT load the webpage body, but only the metadata.

  2. With Open Graph formatting, I want to copy the {og:image} or {og:type} . Is this possible? I can’t get it to work.


after the recent update i cant download images, not touched the setting